@ Pranav Shinde, residual stress in DMLS is a part's "performance" that is a result of the material "response" to the external "stimulus" . Your stimulus is the spatially heterogeneous temporal thermal history during the DMLS process. Your material properties is another spatially heterogeneous time dependent properties. As such, there is no direct answer to your question of the correlation between grain size and residual stresses. Rather, it is a multipscale multiphysics question that can be answered within an integrated computational materials engineering (ICME) effort. This requires a well planned experiments combined with detailed processing-microstructure-performance simulations. The general  experiments you suggested in the DMLS project under your profile need to be combined with other tasks to answer your question. I will need more info about your project to be able to help with answering your question. Please do not hesitate to contact me via a message to discuss that further if you like. Wish you the best of luck! it is a good problem! 

Generally, the hardness is inversely proportional to the square root of grain size. And greater the hardness greater will be the residual stresses. So, we can say that grain sizes have inverse effect on residual stresses. And hardness also depends on type of micro-structure and cooling rate. Like for martensite structure with rapid cooling increase the hardness which leads to increase the residual stresses.

Are there any considerations with respect to intra and inter grain residual stresses?

Hi

You should first consider which form of residual stress you are interested in.

I don't think I agree with the above, grain size offers strengthening but this is different to residual stresses.

Tom

Tom,

considering tensile residual stresses occurring due to thermal misfit in DMLS components.

Pankaj,

What you justified is in accordance with Hall-Petch equation, but the relation only states about yeild stress. So how can we bring it into effect for tensile residual stresses?

I assume that you already know about type-i, type-ii, and type-iii residual stresses. Macrostresses are type-i residual stresses. These are homogeneous over many crystal domains in a material. That is, at the macroscopic scale of the structure. The internal forces and moments associated with macrostresses are equilibrated and balanced on all planes and axes of the volume. It is the Type I residual stresses that have a significant effect on the mechanical performance of materials and, thus, quantitative knowledge of these stresses is useful for manufacturers, operators and regulators of mechanical and structural components and structures, especially in safety critical applications. 

@ Pranav Shinde, residual stress in DMLS is a part's "performance" that is a result of the material "response" to the external "stimulus" . Your stimulus is the spatially heterogeneous temporal thermal history during the DMLS process. Your material properties is another spatially heterogeneous time dependent properties. As such, there is no direct answer to your question of the correlation between grain size and residual stresses. Rather, it is a multipscale multiphysics question that can be answered within an integrated computational materials engineering (ICME) effort. This requires a well planned experiments combined with detailed processing-microstructure-performance simulations. The general  experiments you suggested in the DMLS project under your profile need to be combined with other tasks to answer your question. I will need more info about your project to be able to help with answering your question. Please do not hesitate to contact me via a message to discuss that further if you like. Wish you the best of luck! it is a good problem! 

Dear Pranav Shinde,

Indeed the difference between the types of residual stresses is essential. In terms of X-Ray diffraction, macro-residual stresses may induce line shifts, while the other types of residual stresses may contribute to line broadening. Thermal misfit can induce remarkable line shifts indicating that residual stresses affect the average lattice parameter of the phases concerned, thereby incorporating the effects of all contributing grains. So, indeed macro-residual stresses measured by X-Ray diffraction line shift originate from all contributing grains. An example of the effect of thermal misfit on X-ray diffraction line shift is provided by the attached link. 

Article Unusual lattice parameters in two-phase systems after annealing

1.  On the ORIGINAL question:    residual stresses (RS) are nothing else but elastic accommodation of non-uniform plastic (generally -inelastic, e.g. thermal or phase) strains "in order" to save material continuity. In your (MACRO) case the scales of plastic non-uniformity and hence of resulting RS are comparable to the body dimensions and, evidently, are much greater than a grain size !! On the other hand, of course, the RS due to the interaction of deformed grains or different martensite laths have a typical scale of grain size, whereas in-grain non-uniformity of plastic flow results in sub-grain scales of related RS.

2. There is an INDIRECT influence of grain size (through the yield stress), however, since RS cannot be higher than yield stress.  Thus a plate bent due to RS will somewhat flatten (restore planarity) in heating (decrease of yield stress), insofar as an inevitable plastic accommodation reduces both tensile and compressive stresses in respective layers of the plate.

Is it possible to quantitatively investigate this phenomenon?

MCRO RS can in principle be measured by

1) destructive (cutting out some parts and observing their "back" deformation when free) techniques

and

2)  non-destrictive methods mostly based on diffraction (X-ray or neutron) and evaluating elastic strains (and hence stresses) in UNDER-SURFACE regions.

As to structural scales, they cannot be excavated of course. That is why  approximate indirect methods are used based on ASSUMED correlations between the crystal defectness (curvature) and  some functions of EBSD orientation data.

It should be noted as well that the recently developed HR EBSD (just google it..) can directly measure inhomogeneous  elastic strains within grains and, hence, respective (in-grain) RS due to the grain interaction in deformed polycrystals.

Hi Alexander, "recently" is quite exaggerated. It already exists for 20 years, and I am not really sure how good it is applicable for this problem since it needs quite perfect patterns. And also here we are measuring on a free surface, i.e. not in the bulk (grain).